Thursday Oct 16, 2008

Announcing the AEGIS project - a €12.6m investment in open source accessibility

Today I am more than pleased to share with you news of the AEGIS project, a €12.6m investment in accessibility, with the vast majority of it focused on open source solutions.

What is AEGIS?

AEGIS stands for "open Accessibility Everywhere: Groundwork, Infrastructure, Standards". It is a major research and development investment in building accessibility into future mainstream Information & Communication Technologies.

AEGIS was made possible in large part by the European Commission through a grant under the Seventh Framework Programme funding for Research on eInclusion, in service of the ICT for Inclusion Unit of the Information Society and Media Directorate-General. In addition to the European grant contribution of €8.22m from FP7-ICT-2007-2, matching funds from AEGIS consortium members provide the balance of the €12.6m research and development investment.

Who is AEGIS?

The AEGIS project is made up of a consortium of 20 partners spanning 10 countries. They are drawn from industry (both large & small), disability organizations, and university & commercial research organizations; and they represent much of the leading expertise in accessibility from those sectors. The AEGIS consortium is:

In addition to these main participants, several partners are involving multiple of their departments/branches. At the Center for Research & Technology Hellas, both the Hellenic Institute of Transport and the Informations and Telematics Institute are taking part. At Sun, we are involving our engineering sites in the Czech Republic, Germany, and Ireland in Europe; as well as China and the United States.

When is AEGIS?

AEGIS is a 42 month project that just got underway with a kickoff meeting in Prague last month (posting pictures is on my to-do list...). Right now we are focused on refining our three and a half year plans, planning the initial user studies and requirements gathering, and working on our immediate deliverables (like getting the website up!).

What does AEGIS do?

The AEGIS project objectives are to take the early successes of API-based accessibility solutions (also known as "programmatic" or "engineered" accessibility), and expand upon them in three key areas:

#1 - AEGIS on the desktop

On the open desktop, AEGIS expands upon the existing, good work of the GNOME Accessibility Project, and the Accessibility Project. Key focus areas for the desktop include:

  • magnification improvements (and an open source framework for future magnification-based assistive technologies)
  • ensuring that there is good open source text-to-speech for all European languages
  • ensuring that that is an excellent place for authoring accessible content (and that people with severe cognitive impairments are able to communicate in written form with it)
  • building a robust accessibility testing framework for distributed open source accessibility development
  • developing & integrating real-time-text solutions for the deaf into existing open source audio/video chat software
  • developing open source eye-tracking solutions based on commodity web-cam hardware

#2 - AEGIS work on rich Internet applications ("Web 2.0")

For the web, AEGIS takes the programmatic accessibility ideas already proven in the open desktop, and brings them to the Web 2.0 world of rich Internet applications. Programmatic accessibility is significantly more complex in the web world - even more so with the rich visual interfaces enabled by technologies like AJAX and DHTML and Flash and JavaFX Script.

Instead of the typical three components of the desktop that have to work together - the application (component #1) which exposes accessibility API information as defined by the desktop/operating system (component #2) to assistive technologies (component #3) in order to achieve accessibility - you now have the fourth component of the web browser/web user agent, which sits between the application and the desktop/operating system.

Not only does this web user agent have to convey all of the rich programmatic/API information back and forth between the application and the assistive technology, it also has to play the role of translator - taking some canonical "web accessibility API" and expressing it potentially differently on each operating system it is running on.

In addition to these significant challenges, there is a multiplicity of technologies for creating and deploying rich Internet applications (such as those mentioned above - AJAX, JavaFX Script, etc. - and many others). Emerging specifications from the W3C's Web Accessibility Initiative - the Accessible Rich Internet Applications (WAI-ARIA) - are critical, but aren't by themselves a complete solution.

Web browsers must support these standards, and translate them to the appropriate accessibility interfaces of the underlying operating system, and from there on to assistive technologies. But since web applications are commonly built with user interface component sets, it is critical to build support for these accessibility APIs and standards into those components, so applications built with them don't have to re-implement that support each time.

And to finish the job, developer tools used to create these rich Internet applications should explicitly aid developers in utilizing the accessibility APIs and standards provided by those component sets. And of course, the highly dynamic and visual user interfaces that we see in things like Flash and JavaFX use there own runtimes/user agents, where WAI-ARIA isn't as good a fit.

The AEGIS research and development work for rich Internet applications will be addressing all of these challenges, and doing so for a variety of user agents and user interface component sets on a variety of operating systems. And to the greatest extent possible, doing so utilizing open source technologies.

#3 - AEGIS research into mobile device accessibility

For the past few years, mobile devices have grown in capability and complexity and power such that they now rival yesterday's desktop computers. Yet while most of the desktop computing work is moving to (or in the case of the open source UNIX and GNU/Linux desktop have already moved to) API-based accessibility, what solutions exist for the mobile world today are still stuck in the previous generation approach to accessibility. What assistive technologies exist for mobile are bolt-on, reverse-engineered, and ultimately unsatisfying solutions with limited ability to work with downloaded/3rd party applications - the very place where mobile device capabilities are most rapidly expanding.

In AEGIS we will apply the same proven concepts and approaches of the open desktop to develop research prototypes for accessible mobile devices. Specifically we will work on accessibility frameworks and APIs, user interface component sets, developer tools, and sample applications to enable assistive technologies to work in a supported framework on mobile devices - without the need to reverse-engineer applications.

AEGIS' focus on economic disparities

In the United States where I live, people with disabilities endure a 70% unemployment rate. A 2004 labour market report for Europe notes an average 60% unemployment rate among the blind in Europe - with countries such as Hungary rising to 77% rate. As we move from the West to the Majority World, we starting running into situations such as I commented on recently in Afghanistan, and which Jim Fruchterman so eloquently discusses regularly in his blog. As difficult as it is for unemployed people with disabilities to obtain expensive, reverse-engineered assistive technologies in the West, in much of the rest of the world that isn't even a distant possibility.

We believe that we can dramatically reduce the cost of accessible technology by doing three key things:

  1. shifting from reverse-engineered and bolted on solutions to built-in and supported architectures for accessibility
  2. making the cost of building accessible technology - in developers' time and expense - as close to zero as possible
  3. building a large collection of open source solutions that can be directly used throughout the world with zero licensing or royalty charge

AEGIS' focus on open source

Open source is at the heart of AEGIS. It is the engine that pumps the blood into realizing the widest possible dissemination of the results of AEGIS' research, to enable the broadest possible adoption of AEGIS' techniques and solutions. While not everything done by consortium members under AEGIS will carry an open source license, the vast majority will.

To the greatest extent possible AEGIS work will build on top of existing open source environments and applications and tools and user interface libraries and assistive technologies. This is particularly true of AEGIS' work on the open desktop, but also in AEGIS' work on rich Internet applications.

By leveraging popular (and cross-platform) open source applications like and Firefox we enable large numbers of end users to benefit from our research and development work (and to do so on multiple desktop operating systems).

Likewise by leveraging popular open source developer tools like NetBeans, we enable large numbers of developers to utilize tools for creating accessible applications. And even when the open source technologies aren't part of wildly popular applications that have been downloaded more than 500 million times and are approaching 20% market share, the very fact that so much of AEGIS research and prototypes will be released under open source licenses means that others can incorporate them royalty-free.

And it greatly increases the chances that people with disabilities can get those results at little to no cost. Which bring us to...

AEGIS' focus on people with disabilities

If open source is the heart of AEGIS, then people with disabilities must be its soul. Everything we do in AEGIS starts with them. Some of the strongest disability organizations in Europe are part of AEGIS. Their contributions begin with requirements gathering, continue through their working in partnership with the research & industrial members by participating in development of specifications, and concludes with their critical role in user trials.

It is people with disabilities who have the most to gain from AEGIS research results - and the most to loose if we don't get it right. Technology - so full of promise and potential to help - is at the same time so very capable of leaving them behind. Every second that passes, 4 people are born into this world. In that same second, 36 mobile devices are activated, and 411 web pages are created. Globally we are on track to purchase 1.3 billion mobile devices in 2008. Create 12.8 billion new web pages. Google already indexes 1 trillion web pages! If we don't take care to build accessibility into these now, we run the grave risk of leaving far too many people behind.

AEGIS' focus on developer

It is precisely because of the staggering numbers in the paragraph above that we must also focus on the developer. The easier it is for a developer to create an accessible application or website or mobile device, the more likely it is that she will do so. The cheaper we make it - both in terms of the cost of the tools as well as the cost of developer's time - the more likely we'll see accessible results.

This is why, in addition to having the disability community help us set our requirements and provide input to our specifications and feedback in trials of our prototypes, we will have developers playing exactly those same roles with respect to the developer tools and user interface component sets and accessibility frameworks. Developers from AEGIS consortium industrial partners, and drawn from the computer science departments of University partners, as well as those who join our work in the open source community, will play a vital role in helping us make the process of building accessibility easy and inexpensive.

Sun's role(s) in AEGIS

Sun plays a number of critical roles in the AEGIS project. Sun - in the person of yours truly (and too many long nights) - wrote the bulk of the AEGIS grant proposal. And Sun - again in the person of yours truly - is the Technical Manager of the AEGIS project. But it is other qualities and attributes of Sun Microsystems that will play the biggest role in AEGIS as we carry out this very major undertaking:

#1 - Sun brought the API-based approach to the market

Sun has been the first and most consistent champion of programmatic & built-in accessibility approaches:

  • 1997 with the Java Accessibility API
  • 2000 with StarOffice and accessibility
  • leadership in the GNOME community as maintainers of the GNOME accessibility framework
  • work in the web community as maintainers of the UNIX portion of the Mozilla accessibility project,
  • work on an incredibly powerful and flexible open source screen reader/magnifier

#2 - Sun is the single biggest contributor to open source

According to the UNU-MERIT study in 2006, Sun Microsystems is the single biggest contributor of free and open source software in the world - both by lines of code and by person-months of contribution and by value. Number 2 by ranking isn't even close. In fact, according to this study Sun's contributions are greater than those of thee next 7 largest contributors combined! And with our recent acquisitions of MySQL and Innotek, you'd have to combine the next 10 contributors contributions to exceed Sun's. Since open source is at the heart of so much of AEGIS - including both those places where we have spearheaded much of the existing open source accessibility work and in the less accessible domains of rich Internet applications and mobile devices - Sun's incredible knowledge and expertise an standing within these communities will be critical to AEGIS' open source accessibility success.

#3 - Sun is tied into accessibility standards and regulatory efforts

As we move into a future of converged devices (cell phones that are web browsers and handle our e-mail, Blu-ray movies that include video games and get supplemented with new content all playing on the TV in your living room, and of course the desktop software you can use to make phone calls), there is an increasing push for accessibility regulations that cover these and future emerging technologies. As we move away from software built using a small number of platform-defined programming interfaces of Windows & Macintosh & UNIX, to the "wild west" of anything goes so long as it can send pixels over a the Internet to your screen, there is an obvious need for standards to help define how to make those interfaces accessible.

Sun played a very significant role in the U.S. Section 508 regulatory process which is the framework for technology accessibility in America:

In the standards arena, Sun is a deep and passionate believer in and promoter of open standards that are free to all to implement:

In conclusion...

To close this rather long post, I want to thank the European Commission e-Inclusion unit and my fellow consortium members. With AEGIS we are embarking on a very ambitious and challenging research and development program. I believe it will ultimately be an incredibly satisfying one.


Peter Korn


« July 2016